AINews Portal

News

50 items (filtered)

The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.

SourceAgeTitleSummaryStatus
bluesky
bluesky-ai
2moAnthropic researcher says AI has more than 10% chance of ‘killing all humans’ after colleague quits (over safety concerns) — www.cnbc.com/2…Anthropic researcher says AI has more than 10% chance of ‘killing all humans’ after colleague quits (over safety concerns) — www.cnbc.…
↗Not briefed
arxiv
arxiv-ai
2moContext operations to architecture modelling output from large language models and evaluation criteria for their use in systems engineering designThe development of generative artificial intelligence resources enables opportunities of speeding up systems and engineering design wo…
↗Not briefed
hn
hn-ai
2moSuperAstra – Change SNES games with AI while you play them—
↗Not briefed
bluesky
bluesky-ai
2moIf we want to prevent more autonomous cyberattacks from threatening critical infrastructure, we urgently need stronger regulatory oversight…If we want to prevent more autonomous cyberattacks from threatening critical infrastructure, we urgently need stronger regulatory over…
↗Not briefed
bluesky
bluesky-ai
2moBitte DocGaby's post komplett lesen. Und hier gibt's so Typen, die auf frau (mich) einhacken und mich als unwissend hinstellen, wenn ich d…Bitte DocGaby's post komplett lesen. Und hier gibt's so Typen, die auf frau (mich) einhacken und mich als unwissend hinstellen, wenn…
↗Not briefed
bluesky
bluesky-ai
2moThere’s no way to put the genie back in the bottle. Someone is going to build the next level of AI first, and it might as well be Anthropic…There’s no way to put the genie back in the bottle. Someone is going to build the next level of AI first, and it might as well be Anth…
↗Not briefed
bluesky
bluesky-ai
2moI will be very disappointed if one day I find that these fart smellers are doing this for free, because they could be making a killing from…I will be very disappointed if one day I find that these fart smellers are doing this for free, because they could be making a killing…
↗Not briefed
bluesky
bluesky-ai
2mo“OpenAI didn’t just ignore Michael’s disability—it used it against him. After he disclosed his bipolar diagnosis, the system incorporated t…“OpenAI didn’t just ignore Michael’s disability—it used it against him. After he disclosed his bipolar diagnosis, the system incorpora…
↗Not briefed
rss
techmeme
2moAI sales and marketing startup Clay raised $115M led by Wellington Management at a $7.1B valuation, up from $3.1B when it raised $100M in August 2025 (Michael J. de la Merced/New York Times)Michael J. de la Merced / New York Times: AI sales and marketing startup Clay raised $115M led by Wellington Management at a $7.1B val…
↗Not briefed
bluesky
bluesky-ai
2mo“ChatGPT’s memory feature seemed to build a psychiatric profile, allegedly parsing inputs related to his mental health conditions and then …🔧 ChatGPT“ChatGPT’s memory feature seemed to build a psychiatric profile, allegedly parsing inputs related to his mental health conditions and…
↗Not briefed
bluesky
bluesky-ai
2moRemember there are key distinctions between the LLM companies: -Anthropic: The Culty One -OpenAI: The Deeply Unethical One -X: The Fascist…Remember there are key distinctions between the LLM companies: -Anthropic: The Culty One -OpenAI: The Deeply Unethical One -X: The Fa…
↗Not briefed
arxiv
arxiv-ai
2moSA-Profile: Automated Sulcus Angle Profiling from Super-Resolution MRITrochlear dysplasia (TD) is an abnormality of the femoral trochlea associated with anterior knee pain and patellar instability. The su…
↗Not briefed
bluesky
bluesky-ai
2moI know this example is still cheating, but, man, it’s not really that badI know this example is still cheating, but, man, it’s not really that bad
↗Not briefed
arxiv
arxiv-ai
2moIf It's Not Buggy, Don't Fix It: On the Dynamics of Iterative Bug-fixing with LLMsLarge language models (LLMs) have become ubiquitous in software development, with LLM-based automated program repair tools increasingl…
↗Not briefed
arxiv
arxiv-ai
2moProbPlug: A Plugin Uncertainty Network for Reliable Confidence in LLM Binary ClassificationLarge language models (LLMs) have achieved strong performance across a broad range of classification settings, yet the reliability of…
↗Not briefed
rss
theverge-ai
2moAmazon Prime Video’s new AI tech matches lips to dubbed audioMaxton Hall. | Image: Prime Video Amazon's Prime Video is launching a new AI-powered feature that lines up an actor's mouth with "hu…
↗Not briefed
rss
techcrunch-ai
2moInstacart launches an AI grocery shopping assistant called ClementineInstacart is the latest app to bake a conversational AI assistant into its platform.
↗Not briefed
rss
wired-ai
2moA Stealth Startup Thinks It Just Hacked the Memory ShortageKepler Computing claims a new approach to chip design—and a proprietary material—can help end the supply bottlenecks that have sent me…
↗Not briefed
rss
techcrunch-ai
2moSequoia doubles down on Cymphony as AI agents create new enterprise security risksCymphony was valued at more than $100 million in a $25 million Series A co-led by Sequoia and SMBC Fin Atlas Beyond Fund.
↗Not briefed
bluesky
bluesky-ai
2moThis is somewhat helpful... www.quantamagazine.org/ai-has-solve...This is somewhat helpful... www.quantamagazine.org/ai-has-solve...
↗Not briefed
bluesky
bluesky-ai
2moLe storytelling autour de la sécurité des agents IA a « perdu le contact avec la réalité » OpenAI a confié l’analyse de la cyberattaque d’H…🔧 Hugging FaceLe storytelling autour de la sécurité des agents IA a « perdu le contact avec la réalité » OpenAI a confié l’analyse de la cyberattaqu…
↗Not briefed
hn
hn-ai
2moOpenAI/Hugging Face incident – Independent investigation of agents' behavior [pdf]🔧 Hugging Face—
↗Not briefed
hn
hn-ai
2moALdía – Self-hosted business engine with permissioned MCP tools for agents—
↗Not briefed
hn
hn-ai
2moShow HN: Revliu – Multi-touch attribution from acquisition to revenueI’m a founder and I just built RevliI’m a founder and I just built Revliu.<p>After building several products, I always wanted to under…
↗Not briefed
bluesky
bluesky-ai
2moit's more insidious than even just profit seeking. AI is the wall that will eventually shut out the underclasses. just forever screaming at…it's more insidious than even just profit seeking. AI is the wall that will eventually shut out the underclasses. just forever screami…
↗Not briefed
bluesky
bluesky-ai
2moLLM engineers be likeLLM engineers be like
↗Not briefed
hn
hn-ai
2moPost-Training RLM Agents for End-to-End M&A Diligence—
↗Not briefed
bluesky
bluesky-ai
2moAn obvious but boring way to prevent this would be to not give AI access to anything that could kill people. Alas companies like Anthropic …An obvious but boring way to prevent this would be to not give AI access to anything that could kill people. Alas companies like Anthr…
↗Not briefed
bluesky
bluesky-ai
2moI mean, this is a foundational belief for the cultists that run Anthropic. It has been since before the created Anthropic. It is roughly e…I mean, this is a foundational belief for the cultists that run Anthropic. It has been since before the created Anthropic. It is roug…
↗Not briefed
hn
hn-ai
2moShow HN: DortDB, a TS library for mixed-lang queries on JavaScript structuresHello HN,<p>I made a TypeScript library for querying in-memory JS data. Originally, the goal was to have a more modular and lightweigh…
↗Not briefed
hn
hn-ai
2moDraw on Mapbox, MapLibre, Google, Leaflet, or ArcGIS in React—
↗Not briefed
bluesky
bluesky-ai
2moYou could argue people using the AI to write bloated code that only gets used once or twice is a waste of power and I’d agree, but let’s no…You could argue people using the AI to write bloated code that only gets used once or twice is a waste of power and I’d agree, but let…
↗Not briefed
hn
hn-ai
2moShow HN: Seams handles access, spend caps, and metering on one request pathSeams is the billing infrastructure for AI products, we help you manage end user usage spend for different pricing strategies, usage-b…
↗Not briefed
bluesky
bluesky-ai
2moWhatever the truth of it, the idea that AI risks destroying everything is an idea its own luminaries promote.Whatever the truth of it, the idea that AI risks destroying everything is an idea its own luminaries promote.
↗Not briefed
bluesky
bluesky-ai
2moRe: Anthropic pre-crime story going aroundRe: Anthropic pre-crime story going around
↗Not briefed
hn
hn-ai
2moShow HN: GuardRail, shell guards that stop Claude Code before it pushes to main—
↗Not briefed
bluesky
bluesky-ai
2moAnthropic S-1 RISK FACTORS - We have a limited operating history, which makes it difficult to forecast our revenue and future prospects …Anthropic S-1 RISK FACTORS - We have a limited operating history, which makes it difficult to forecast our revenue and future prospe…
↗Not briefed
bluesky
bluesky-ai
2mogringos "redescobrindo" espiritualidade pra justificar meter llm em tudo 🙄gringos "redescobrindo" espiritualidade pra justificar meter llm em tudo 🙄
↗Not briefed
rss
techmeme
2moOpenAI says it is working with Samsung on its next-gen chips, and Samsung is one of the "largest-scale deployments of ChatGPT globally" (Heekyong Yang/Reuters)🔧 ChatGPTHeekyong Yang / Reuters: OpenAI says it is working with Samsung on its next-gen chips, and Samsung is one of the “largest-scale deploy…
↗Not briefed
hn
hn-ai
2moShow HN: I rewrite my game while it is still running—
↗Not briefed
rss
the-decoder
2moAnthropic scientist puts the odds of AI destroying humanity above ten percent this decadeJacob Coxon, a former pretraining researcher at OpenAI and Anthropic, has quit and accuses both companies of knowingly risking human e…
↗Not briefed
bluesky
bluesky-ai
2moIt's important to remember the origin story for a good fraction of the LLM makers:It's important to remember the origin story for a good fraction of the LLM makers:
↗Not briefed
hn
hn-ai
2movLLM 0.29.0: Model Runner V2 is now the default🔧 vLLM—
↗Not briefed
bluesky
bluesky-ai
2moPotentially Misanthropic, then. www.bbc.com/news/article...Potentially Misanthropic, then. www.bbc.com/news/article...
↗Not briefed
hn
hn-ai
2moShow HN: GPT 6 Astra in Lanes—
↗Not briefed
arxiv
arxiv-ai
2moData-Centric Post-Training for Financial Reasoning: Mining, Distillation, and Verifiable LearningFinancial text, textbooks, and question-answer pairs are abundant, but only a small fraction is directly usable for reasoning-focused…
↗Not briefed
bluesky
bluesky-ai
2moThis isn’t mentioned a lot. OpenAI can have several civilizations of agents running wild because they aren’t paying for tokens. I worked i…This isn’t mentioned a lot. OpenAI can have several civilizations of agents running wild because they aren’t paying for tokens. I wor…
↗Not briefed
arxiv
arxiv-ai
2moStorage-Scalable Progressive Semantic Communication via Knowledge-Base ReuseExisting knowledge-base-assisted semantic communication schemes commonly adopt either single knowledge-base quantization (SKBQ) or mul…
↗Not briefed
hn
hn-ai
2moIt's Getting Harder to Return Things–and It's Making Shoppers Mad—
↗Not briefed
bluesky
bluesky-ai
2moRicercatore di Anthropic si dimette: 'L'umanità a rischio entro la fine del decennio'Ricercatore di Anthropic si dimette: 'L'umanità a rischio entro la fine del decennio'
↗Not briefed